Fixed mapping for Minkowski-P distance normalisation
B = NORMM(A,P)
Normalizes the distances of all objects in the dataset A such that their Minkowski-P distances to the origin equal one. For P=1 (default), this is useful for normalizing the probabilities. For 1-dimensional datasets (SIZE(A,2)=1), a second feature is added before the normalisation such that A(:,2)=1-A(:,1).
Note that NORMM(P) or NORMM(,P) is a fixed mapping.